Too many made up numbers.
The only ones which should matter are the ones which vary, so that'd be Haste and Accuracy. Everything else should be symbolic...
Damage/minute = dmg/swing *
swing/min *
accuracy (Setup 1) Haste (
h1) = 20%; Accuracy (
acc1) = 63%
(Setup 2) Haste (
h2) = 23%; Accuracy (
acc2) = 60%
(Setup 1) swing/min (
spm1) = 3600 / (delay * (1 - h1)) = 3600/
D1 (Setup 2) swing/min (
spm2) = 3600 / (delay * (1 - h2)) = 3600/
D2 D1 = delay * (1 - h1)
D2 = delay * (1 - h2)
(Setup 1) Damage/minute (
dpm1) = (dmg/swing) * (3600/D1) * acc1
(Setup 2) Damage/minute (
dpm2) = (dmg/swing) * (3600/D2) * acc2
To Compare:
dpm1/
dpm2 = ((1/D1) * acc1) / ((1/D2) * acc2)
= D2/D1 * acc1/acc2
= (1 -
h2)/(1 -
h1) * (
acc1 /
acc2)
= (1 - 23%)/(1 - 20%) * (63%/60%)
= 1.0106
Conclusion:
(Setup 1)
is better than (Setup 2); or,
(20% Haste, 63% Accuracy)
is better than (23% Haste, 60% Accuracy).
* * *
If I did everything right, to compare just accuracy and delay of two different setups, it'd be just:
(
Setup 1)/(
Setup 2)
= (1 -
h2)/(1 -
h1) * (
acc1 /
acc2)
* * *
Incidentally, with more realistic accuracy numbers for sushi eaters, such as:
h1 = 20%; acc1 = 93%
h2 = 23%; acc2 = 90%
dpm1/dpm2 = 0.9945
Setup 2 would win the damage over time game.