If
you can calculate how a feat, a weapon, a spell, a tactic, or
a magic feature affects your performance in combat, you can better
select those options that help you most.
Average
Damage per Round
A
key statistic to know is your average (mean) damage per round.
Basically, you multiply the number of successful hits you get
in the average round by the amount of damage you deal per hit.
(Since the number of hits you can expect per round varies by the
opponents AC, you need to look at different ACs over a range
of typical ACs.)
Example:
At 1st level, Tordek the dwarven fighter attacks at +4 for 1d10+2
damage. His average damage vs. AC 16 (an ogre) is 45% (chance
to roll 12+ on 1d20) times 7.5 (average damage) for a total of
3.375. His average damage vs. AC 11 (a zombie) is 70% ´ 7.5 =
5.25.
Warning:
When you calculate average damage this way, you assume that every
point of damage you deal counts. Against a kobold with 2 hit points,
Tordeks average damage with a successful hit is 3 points.
Any more damage than that is wasted. Thus, he deals the same average
damage per hit with his fist as with a dwarven waraxe. When youre
fighting creatures with fewer hit points than you can deal damage,
not all of your damage capacity "counts."
Criticals
and Average Damage
Think
of a standard critical as "a 5% chance to deal +100% damage,"
which averages out to +5% damage.
A
swordlike critical (threat on 19 and 20) is "a 10% chance
to deal +100% damage," which averages out to +10% damage.
An
axelike critical (triple damage) is "a 5% chance to deal
+200% damage," which also averages out to +10% damage.
Thus,
Tordeks average damage is actually 3.7125 per round vs.
ogres. His triple criticals dont improve his damage vs.
zombies because undead are not subject to critical hits.
Attack
Stats
Sometimes
you need to choose between more attacks, higher attack bonuses,
and more damage. For instance, a monk can use a "flurry of
blows" to get one extra attack at the cost of suffering 2
on all attack rolls. The Power Attack feat also lets you decrease
your attack bonus to increase your damage. We can compare how
different benefits affect average damage.
Example:
Ember, a 1st-level monk, is fighting a monster with AC 14 (such
as an orc). She can strike with one unarmed attack (attack +2)
or with two, as a flurry (attack +0). Her number of expected hits
per round is .45 with one attack (her chance to roll 12+ on 1d20),
and her number of expected hits per round is .7 with two attacks
(35% chance to roll 14+, twice). Wrinkle: If Ember is fighting
an orc, theres a 23% chance that she wont have an
opportunity for a second attack (because the first attack does
the orc in). Thus, her chance to hit with the second attack is
really only 77% ´ 35%, or 27%. Thus her expected number of hits
against the orc is .35 + .27 or .62 (still better than .45 with
one attack.)
Example:
At 1st level, the human fighter Regdar (attack +4, damage 11 with
a two-handed sword [damage 10 vs. creatures not subject to criticals])
has Power Attack. Against low AC creatures, its a good idea
to use it. Against high AC creatures, its a bad idea. This
table shows Regdars average expected damage against these
creatures.
| |
Gelatinous
Cube (AC 3)
|
Zombie
(AC 11)
|
Ogre
(AC 16)
|
Ogre
in Half
Plate (AC 20)
|
| Normal
Attack |
9.5
|
7.0
|
5.0
|
2.8
|
| Power
Attack (1/+1) |
10.5
|
7.2
|
4.8
|
2.4
|
Randomness
vs. Predictability.
Raw
calculations of averages doesnt take into account randomness.
Player characters want to reduce randomness. Randomness
favors the underdog, and PCs are usually the favored side in a
fight. Thus, the more randomness there is in a fight, the worse
it is for the PCs.
Hypothetical:
Imagine a monster that dealt 1 damage to one character every round
and another that had a 1 in 20 chance to deal 20 damage to one
character every round. Their average damage would be the same,
but the one capable of getting lucky and dealing 20 damage all
at once is much more likely to kill a 1st-level PC.
Orcs:
Orcs have only about 4 hit points, and their AC is only 14, but
they wield greataxes that deal 1d12+3 damage with triple damage
criticals. They dont last long, and their average damage
isnt that bad, but if you fight orcs several times, one
of them is bound to get lucky eventually and hit a character with
a critical (average damage 28.5).
Offense
vs. Defense
Sometimes
it pays to improve your attack at the expense of your defense
(such as by charging). Sometimes it pays to improve your defense
at the expense of your attack (such as with Expertise).
Example:
Tordeks facing a goblin. Should Tordek charge (+2 attack,
2 AC) or attack normally?
|
Stats
|
|
Tordek:
+4 attack, AC 17, d10+2 damage
|
|
Goblin:
+0 attack, AC 15, hp 5
|
|
Combat
|
|
Tordek
chance to drop, normal: chance to hit ´
chance to deal 5 damage.
|
|
|
=
50% x
80% = 40%
|
|
Tordek
chance to drop, charge:
|
|
|
=
60% x
80% = 48%
|
|
Goblin
chance to hit Tordek, standard:
|
|
|
=
20%
|
|
Goblin
chance to hit Tordek, charge:
|
|
|
=
30%
|
|
Chance
for goblin to survive Tordeks attack and hit him.
|
|
|
If
Tordek attacks normally: 60% x
20% = 12%
|
|
|
If
Tordek charges: 52% x
30% = 15.6%
|
|
Tordek
is fighting the goblin. Should he fight normally or defensively
(4 attack, +2 AC)?
|
|
|
Tordeks
chance to hit the goblin, normal/defensive: 50%/30%
|
|
|
Goblins
chance to hit Tordek, normal/defensive: 20%/10%
|
|
|
Fighting
normally, Tordek hits the goblin two and one-half times
as often as the goblin hits him. Fighting defensively, Tordek
hits the goblin three times as often as the goblin hits
him.
|
|
|
Fighting
defensively has a hidden benefit. The longer a fighter can
drag out a combat, the more spells the spellcasters will
have time to cast.
|
About
the Author
Jonathan
Tweet, a senior designer at Wizards of the Coast, led the 3rd
Edition D&D design team and authored the new Players
Handbook. Since 1986, he has freelanced, self-published, and
worked full time in the adventure game industry. His other design
credits include Ars Magica, Over the Edge, Everway, and
support material for AD&D, Magic: The Gathering,
Netrunner, RuneQuest, and Talislanta.