Quote:
Originally Posted by Duik
[You must be logged in to view images. Log in or Register.]
Would constantly swapping 1hb to 2hb mitigate the highs/lows that are inherant in any randomized data set? !
|
Since you're one of my favorite posters on this site, I spent some time to explore this question. Short answer is no. All that matters in terms of highs/lows away from the expected DPS is the total number of swings. The more times you swing, the more likely that the actual DPS is "close" to the expected/average DPS.
All that swapping between 1h and 2h does is give you a result somewhere between the result with only-1h and only-2h.
Let's say you have a 1H weapon set and a 2H weapon that both have the same expected DPS. Let's say the delays are such that you'd have half as many swings with 2h as 1H per minute. Let's say you're soloing an evenly matched mob such that all 20 values of the damage interval are equally likely. Let's say that you're confident you'll win the fight as long as you do at least half the expected DPS. Let's say the fight lasts long enough that there's an 99.94% chance of doing between half and twice the expected DPS with the 1handers. In that scenario, the 2-hander will have a 68% chance of doing between half and twice the expected DPS. Of the remaining 32%, 16% of the time you'll do less than half the expected DPS, and 16% of the time you'll do more than twice the expected DPS.
So with the 1-handers there's a 99.97% chance you'll survive the fight and an 84% chance with the 2-hander. Not taking into account any variance in the mob's damage, of course.
Numbers are taken from standard deviations of a bell curve. I'm assuming 1 standard deviation with the 1-handers and 2 standard deviations with the 2-hander, which is supported by the law of large numbers and the central limit theorem. The central limit theorem says that the variance of n samples from a distribution with variance V is V/n: double the samples, halve the variance. Variance is the square of the standard deviation, so I'm comparing 4 standard deviations for 1-handers to 1 standard deviation for 2-handers.
I'd love to get 7thGate's thoughts on this, as I'm not super confident in how I'm using standard deviations here. I've never thought about how different variances between 1- and 2-handers would affect survivability in a solo fight. And thanks to DSM for raising such an interesting topic, even if it has next to no impact on a raid encounter that requires 10-20,000 hits to kill.