Iterative Attacks

[Gantros]

In general, I like the original idea. I think it's an improvement over the standard iterative attack rules in terms of speed and simplicity. On the other hand, it still has the downsides of reducing the number of attacks a high-level fighter can make, and it still requires 4 rolls (2 attacks + 2 damage) per full attack with all of their attendant modifiers.

Here's an alternative suggestion I came up with. Basically you can choose from two ways to use iterative attacks - against a single opponent, or against multiple opponents.

Single opponent

Make a standard attack. If it misses, no further attacks are possible against that opponent this round. If it hits, roll damage and apply any modifiers or critical effects.

At 6th level or higher, if the attack beats the target AC by 5 or more, total damage is doubled.
At 11th level or higher, if the attack beats the target AC by 10 or more, total damage is tripled.
At 16th level or higher, if the attack beats the target AC by 15 or more, total damage is quadrupled.

Multiple opponents

Make separate attack and damage rolls for each opponent.

At 6th level, you can attack up to two opponents, each at -2 to hit.
At 11th level, you can attack up to three opponents, each at -4 to hit.
At 16th level, you can attack up to four opponents, each at -6 to hit.

Why do it this way? In my experience, iterative attacks are most commonly used to beat on a single opponent, and in this case it seems unnecessary to require multiple rolls with different modifiers. This method lets you get away with one attack roll and one damage roll in all such cases, at the cost of a little bit of extra math. It also gives you the option of hitting multiple opponents as before, which still requires multiple rolls and bonus calculations, but it's simplified by having a single modifier for all attacks. The modifiers were selected to make the 3rd and 4th attacks more meaningful, while keeping the average probabilities of scoring hits relatively unchanged.

You could potentially even combine the single + multiple attack options if desired (e.g. a 16th level fighter attacks two opponents at -2 each, with 2x damage if he beats either AC by 5 or more).

[kibbitz]

Originally Posted by Gantros

In general, I like the original idea. I think it's an improvement over the standard iterative attack rules in terms of speed and simplicity. On the other hand, it still has the downsides of reducing the number of attacks a high-level fighter can make, and it still requires 4 rolls (2 attacks + 2 damage) per full attack with all of their attendant modifiers.

Here's an alternative suggestion I came up with. Basically you can choose from two ways to use iterative attacks - against a single opponent, or against multiple opponents.

---snip---

You could potentially even combine the single + multiple attack options if desired (e.g. a 16th level fighter attacks two opponents at -2 each, with 2x damage if he beats either AC by 5 or more).

Interesting idea, but the need to beat the AC by a certain amount conflicts with the existing mechanic for Power Attack. If you went this route, how would you remodel Power Attack?

[Wulf Ratbane]

Gantros, what are the damage expectations on that method?

[Gantros]

Originally Posted by kibbitz

Interesting idea, but the need to beat the AC by a certain amount conflicts with the existing mechanic for Power Attack. If you went this route, how would you remodel Power Attack?

I haven't worked out all the math for it, but I think you should be able to use Power Attack normally. You would just be trading a reduced chance of getting a damage multiplier for an increase in base damage.

Example: Joe the 13th level fighter has +21 to hit vs. AC20 opponent. Normally he'd hit on a 2 or better, do 2x damage on a 4 or better, and 3x damage on a 9 or better. He opts for a +5 power attack, so now he hits on a 4 or better, gets +5 to damage before multipliers, 2x damage on a 9 or better, and 3x damage on a 14 or better.

Originally Posted by Wulf Ratbane

Gantros, what are the damage expectations on that method?

By my calculations, the single opponent method results in equal to slightly less damage than the RAW. The delta ranges from 0 to -15%, with the biggest difference when using 4 attacks vs. high AC opponents. The variance is due to the fact that iterative attacks can always still hit on a natural 20.

[Wulf Ratbane]

Originally Posted by Gantros

By my calculations, the single opponent method results in equal to slightly less damage than the RAW. The delta ranges from 0 to -15%, with the biggest difference when using 4 attacks vs. high AC opponents. The variance is due to the fact that iterative attacks can always still hit on a natural 20.

I considered a single roll with "levels of success" early on.

I would expect just glancing at it that you'll chop off the top half of the edge cases (high AC) and underperform against the bottom half (low AC). That would be fine as long as there's some compensation in the middle.

EDIT: For the record, I threw out the levels of success model because it requires extra math after the roll to calculate the result.

[Gantros]

Actually this method eliminates the low AC problem - in those cases average damage is identical to the RAW, since all of your attacks were almost certain to hit anyway.

The variance only becomes significant when you have >2 attacks and a low probability of scoring a hit on your initial attack, which is a relatively uncommon situation for high level fighters. Essentially, you're only missing out on those occasions where you'd get lucky and roll a natural 20 on one of your extra attacks, which the damage multiplier doesn't account for. Personally I think those occasions, when the fighter manages to deal big damage to a high AC opponent, are rare enough that the speedup in all other situations more than compensates for their loss.

As far as extra math goes, I can't really see a situation where this would require any more of it than your proposal. In the worst case, it requires 1 modified attack roll + 1 AC comparison + 1 modified damage roll + 1 damage multiplier. Your method would require 2 modified attack rolls + 2 AC comparisons + 2 modified damage rolls. The most time consuming part is adding up the modifiers for each roll (plus crit damage, and variable magic weapon effects), so reducing the number of rolls whenever possible is what saves you time and math.

[Wulf Ratbane]

Originally Posted by Gantros

As far as extra math goes, I can't really see a situation where this would require any more of it than your proposal. In the worst case, it requires 1 modified attack roll + 1 AC comparison + 1 modified damage roll + 1 damage multiplier. Your method would require 2 modified attack rolls + 2 AC comparisons + 2 modified damage rolls.

No. Your method requires comparing the attack roll to the AC, then figuring out how much you beat it by, in increments of 5. That is not err, is a non-zero calculation time.

[Wulf Ratbane]

I've run the levels of success method and see that it lags behind 0/5/10/15 (4 attacks, RAW) between .09 and .19 at every possible step-- for an average expectation loss of .15.

Definitely not down with that.

[whydirt]

What if you made the margins of success +4/+8/+12 instead?

[Gantros]

Maybe I'm missing something, but I come up with an average loss of 7.5% across the entire to-hit range with 4 attacks, which is half of what you're getting.

Here's how I calculated:

Let's take a case where you need to roll a 5 or better to hit. By the RAW, that's an 80% hit rate for the 1st attack, 55% for the 2nd, 30% for the 3rd, and 5% for the 4th. Since each attack is roll is independent, expected damage per round is the sum of these, or 170% of the average damage for a single attack.

With level of success, there is a 20% chance for a miss, a 25% chance for normal damage, a 25% chance for 2x damage, a 25% chance for 3x damage, and a 5% chance for 4x damage. Adding this up gives an expected damage identical to that of the RAW.

Using similar calculations for the worst case, when a roll of 16 or higher is required to hit with 4 iterative attacks, the RAW gives an expected damage of 0.4 while level of success gives 0.25.

[pogre]

As a DM I like it.

I ran several campaigns into the epic levels and iterative attacks annoyed me.

I did have one player who loved rolling handfuls of dice though. He played a high level rogue and relished rolling a few d20's and bucketloads of d6's. I'll have to ask him what he thinks.

He really, really wants to go back to 3.X and I cannot help but think that bucketloads of dice are part of the reason.

[Wulf Ratbane]

Originally Posted by Gantros

Maybe I'm missing something, but I come up with an average loss of 7.5% across the entire to-hit range with 4 attacks, which is half of what you're getting.

It looks like we're using the same methodology-- but I wonder if you're stopping your chart earlier?

I started from the last attack and worked backwards-- cases where you can successfully hit "By 15 or more" on a natural 2.

My chart runs from AC-attack bonus = -13 to +20.

I am perfectly capable of user error, I'll tell you that.

(It's certainly curious that you're getting exactly half what I am getting.)

EDIT 2: Found it. I'm also counting x1.05 for crits. Backing that out, and shortening the chart down to just natural 2 on the first attack, accounts for the difference. We also have a slightly different method for expecation-- I am using long form permutations, but that accounts for a difference in the hundredths of points.

So to reiterate: Your methodology looks fine. An across the board drop in damage won't work for me, but if it works for you, it seems at least you're thinking it through solid.

[Gantros]

Thanks for checking Wulf, glad to hear my stats skills haven't completely atrophied yet I was only averaging from 0 to +20, so that explains our different averages. If I extend the range to -13, the average variance drops to -4.4%.

By the way, if you want the level of success method to exactly match the expected damage from the RAW in all cases, there's a simple fix - just add the rule that if you roll a natural 20, you always get the max damage multiplier, regardless of how much you beat the AC by.

[Kid Charlemagne]

Originally Posted by pogre

I did have one player who loved rolling handfuls of dice though. He played a high level rogue and relished rolling a few d20's and bucketloads of d6's. I'll have to ask him what he thinks.

He really, really wants to go back to 3.X and I cannot help but think that bucketloads of dice are part of the reason.

Bucket loads of dice are fun - especially for damage (soem people are hopeless when it comes to adding them up, though). For a rogue, they won't get as many iterative attacks so I wouldn't think it would be as big of a deal as it is for a fighter...

[roguerouge]

Since the existence of scrap paper and the concept of pre-rolling eliminates many of the math concerns in my campaigns that so many people are on about, my only problem with iterative attacks is the loss of a move action to use them. I wouldn't change how iterative attacks work except for that.

[Wulf Ratbane]

Originally Posted by Gantros

Thanks for checking Wulf, glad to hear my stats skills haven't completely atrophied yet

If you have stats skills, you're ahead of me. I keep "Probability For Dummies" near at hand.

I actually used the "long form" probability for four attacks, just to be thorough and make sure I had set up the probability problem correctly. You might chuckle to see my spreadsheet-- for a four attack sequence, I calculate each of the 16 possible combinations (1-4-6-4-1) to double check that it occupies the correct proportion of the 160,000 total outcomes.

Yeah, long form.

By the way, if you want the level of success method to exactly match the expected damage from the RAW in all cases, there's a simple fix - just add the rule that if you roll a natural 20, you always get the max damage multiplier, regardless of how much you beat the AC by.

Cool, I'll check that. I wouldn't mind putting that option up against others in a poll. That's very useful.

Walk me through that in the case of criticals.

If you roll a natural 20, you automatically get the full "iterative attack" multiplier. You also score a critical threat-- does the confirmation roll also enjoy the same benefits on a natural 20? In other words, does a 20 followed by a 20 score the full iterative multiplier, with the "base damage" equal to your critical damage roll?

I think your expectation might fall off against weapons with wider threat ranges (19-20, 18-20) that benefit from multiple rolls.

Nice contribution to the thread overall; I am much obliged.

[Wulf Ratbane]

Originally Posted by roguerouge

Since the existence of scrap paper and the concept of pre-rolling eliminates many of the math concerns in my campaigns that so many people are on about, my only problem with iterative attacks is the loss of a move action to use them. I wouldn't change how iterative attacks work except for that.

I looked at the idea of allowing multiple attacks even while moving. The problem with that is that it actually punishes the PCs in the long term. They fight a lot more monsters with multiple attacks, often a lot earlier than they themselves have multiple attacks.

The ability to keep on the move and not stand toe to toe with a troll is pretty important in mitigating the effect of its Rend. Same for giants, same for dragons. Mobility helps the PCs more, I think.

[roguerouge]

Gantros: my DMs typically don't tell me the AC of what I'm attacking. So doesn't this undermine the time-saving practice of pre-rolling? In your system, I have to ask if each attack hits and by how much. Once I know that, I have to multiply the damage to that hit and write that number down. I do this for each of the four attacks and then sum all that damage that's been multiplied (or not). In the RAW way, you pre-roll your attacks and damage. Then you ask your DM, "Does a 36 hit? No? Does a 39 hit? Yes!" Then I add all the damage from the attacks that hit AC 39 or better and announce the damage and the number of hits. I feel like the multiplication under your system would slow things down at high levels just a bit.

[roguerouge]

Originally Posted by Wulf Ratbane

I looked at the idea of allowing multiple attacks even while moving. The problem with that is that it actually punishes the PCs in the long term. They fight a lot more monsters with multiple attacks, often a lot earlier than they themselves have multiple attacks.

The ability to keep on the move and not stand toe to toe with a troll is pretty important in mitigating the effect of its Rend. Same for giants, same for dragons. Mobility helps the PCs more, I think.

Shoot. You're right. Maybe that's a good feat to design then, which at least limits the access to monsters.

[Wulf Ratbane]

Originally Posted by roguerouge

Gantros: my DMs typically don't tell me the AC of what I'm attacking. So doesn't this undermine the time-saving practice of pre-rolling? In your system, I have to ask if each attack hits and by how much.

I have my issues with Gantros' solution (though they are slowly eroding) but I am going to back him up here-- tangentially.

One of the things I mention in in my advice for speeding up play is to dispense with the "AC is secret!" mentality.

My players are smarter than the average bear-- well, mostly, except for the fact that they can't add up bonuses quick enough to suit me. But they add them up between rounds, and they keep notes, and so within about 3 rounds they have the monster's AC pegged to within a couple of points, anyway.

(EDIT: Tangential note. I expect my players to do this-- what I did not expect was that the ringleader behind these careful meta-gamey notes was the LARP'er chick at the table. We're going to miss her, and her "I rage, and I shift, and I power attack for everything I can!"-- graduate school stole her away from us. )

Just dispense with the false suspense: Tell the PCs what the AC is.

The gain you get in a smooth running game far outstrips the benefits of "secret AC."

Originally Posted by roguerouge

Shoot. You're right. Maybe that's a good feat to design then, which at least limits the access to monsters.

Two very quick comments, admittedly without a lot of thought behind them:

1) I have no problem with PCs being able to do things the monsters can't, and vice versa.

2) There are existing feats that are similar-- Superior Cleave (Oriental Adventures) lets you take a 5 foot step after you drop an opponent. I could certainly get behind a similar feat-- or, frankly, a blanket permission-- that allows a fighter to move after dropping an opponent, provided he has not moved yet. Basically, allow you to break up your full attack at any point, with one move, provided that such movement wouldn't otherwise provoke an AoO.

But not attack -move a bit - attack -move a bit more - attack, etc.

Hmm... The levels of success method breaks such a solution. Gantros, do you allow the attack multiplier if the combatant has moved?

And while I am thinking of it, do you do anything different with multiple attacks? (Claw/Claw/Bite routines and the like.)