2012-04-09, 04:50 AM
The incremental damage increase for %total vs. %crit looks like this.
![[Image: Uqvsm.png]](http://i.imgur.com/Uqvsm.png)
What this means is that, if you have 0% total damage, then 1% more is equal to 1% more damage (top left). If you have 100% total damage, 1% more is equal to only 0.5% more damage (mid right).
The reason % crit is so much lower is, simply - 100% crit only adds ~35% damage.
In addition, most classes start with some base crit percent - also shifting the value to the right, making more %crit less valuable. It's not as extreme as the %total because it's not adding as much damage at any point. The final data point is 0.26% damage increase from 1% crit, whereas the first is 0.35%.
So when you add 8% crit, you're basically summing 8 sequential points on this plot - eg sum(crit[45 to 52]), and when you add 4% total you're summing 4 sequential points, sum(total[1 to 4]).
Actually doing those two specifics, you get 2.4% damage increase from 8% crit, vs. 4% damage increase from 4% total damage.
If you want the plot for yourself, R code looks something like
You can change the .35 in the crit to reflect a particular character's (min+max)/2 crit damage. Default is 120~150% = 135% or 0.35
For the posted numbers (50% crit + 135~165%) with +6% crit Nebulite, the incremental improvement is 2.4% from the crit nebulite. (someone correct me if I'm wrong about Decent SE being 15% max) which is still worse than 4% total damage.
edit: methodology on sums is slightly off. You want (total[b]-total[a])/total[a], of course.
![[Image: Uqvsm.png]](http://i.imgur.com/Uqvsm.png)
What this means is that, if you have 0% total damage, then 1% more is equal to 1% more damage (top left). If you have 100% total damage, 1% more is equal to only 0.5% more damage (mid right).
The reason % crit is so much lower is, simply - 100% crit only adds ~35% damage.
In addition, most classes start with some base crit percent - also shifting the value to the right, making more %crit less valuable. It's not as extreme as the %total because it's not adding as much damage at any point. The final data point is 0.26% damage increase from 1% crit, whereas the first is 0.35%.
So when you add 8% crit, you're basically summing 8 sequential points on this plot - eg sum(crit[45 to 52]), and when you add 4% total you're summing 4 sequential points, sum(total[1 to 4]).
Actually doing those two specifics, you get 2.4% damage increase from 8% crit, vs. 4% damage increase from 4% total damage.
If you want the plot for yourself, R code looks something like
Code:
x = 1:100
total = 1+x/100
difftotal = (total[2:100]-total[1:99])/total[1:99]
crit = 1+.35*x/100
diffcrit = (crit[2:100]-crit[1:99])/crit[1:99]
plot(difftotal*100,type='l',ylim=c(0,1),xlab="Percent bonus",ylab="Incremental improvement from 1% more")
lines(diffcrit*100)
(crit[52]-crit[44])/crit[44]
(total[4]-1)/1For the posted numbers (50% crit + 135~165%) with +6% crit Nebulite, the incremental improvement is 2.4% from the crit nebulite. (someone correct me if I'm wrong about Decent SE being 15% max) which is still worse than 4% total damage.
edit: methodology on sums is slightly off. You want (total[b]-total[a])/total[a], of course.

