So, let me start off by saying that I'm not claiming Relentless Attack/Unforgiving Chain are terrible because I have 100% crit chance from Cunning promotions (I don't). Nor am I even saying they're terrible because I have 60% crit or 40% crit or whatever number you want to throw in.
They're still terrible even at the base 5% crit where they give an average bonus of 6.3-6.4%. And if you assume the character has at least some other crit boosters (from gear, promotions, or class skills) to get to even 15% crit...that bonus drops to 3.8% overall crit.
To put that in perspective, most people think the passives that give pure stats are pretty bad. And those passives give 9 total stats. 9 Cunning alone is 4.5% crit *and* 4.5% reduced ranged damage...and half of that bonus is already better than the crit chain talents with 15% starting crit.
Yeah.
So here are some various numbers for perspective (all of these were generated using 100,000 trials and the code is included at the end of the post for people who want to test it themselves):
5% starting crit, 2% crit increase: 6.32% crit from talent
15% starting crit, 2% crit increase: 3.76% crit from talent
30% starting crit, 2% crit increase: 1.99% crit from talent
50% starting crit, 2% crit increase: 0.93% crit from talent
75% starting crit, 2% crit increase: 0.32% crit from talent
5% starting crit, 4% crit increase: 9.45% crit from talent
15% starting crit, 4% crit increase: 6.27% crit from talent
30% starting crit, 4% crit increase: 3.56% crit from talent
50% starting crit, 4% crit increase: 1.77% crit from talent
75% starting crit, 4% crit increase: 0.63% crit from talent
5% starting crit, 10% crit increase: 15.58% crit from talent
15% starting crit, 10% crit increase: 11.42% crit from talent
30% starting crit, 10% crit increase: 7.22% crit from talent
50% starting crit, 10% crit increase: 3.88% crit from talent
75% starting crit, 10% crit increase: 1.46% crit from talent
So...yeah. I'd also point out that given Rogues get Cunning from quite a few of their skills in the first place it's basically impossible for them to really even be at 5% crit. Which means even a 10% crit increase per attack seems reasonable as it's likely to be at 12.5% overall increase best case and for people with reasonable gear/levels (even ignoring promotions) I wouldn't be surprised if it was usually 8% crit or less.
Maybe I wildly screwed up the following C code (I almost hope so, I slapped it together quickly) but some basic tests seemed to pan out correctly...
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
float ChainFunction(float fCrit, float fCritIncrease)
{
float fTotalCrit = fCrit, fRandom;
int nPass = 1;
while(1)
{
fRandom = ((float) rand()/(float)(RAND_MAX))* 100.0;
if (fRandom < fCrit)
{
return fTotalCrit/nPass;
}
fCrit += fCritIncrease;
fTotalCrit += fCrit;
++nPass;
}
}
int main(void)
{
srand(time(NULL));
float fCrit, fCritIncrease, fReturn, fCritSum = 0.0;
int nTrials, nCounter = 1;
printf("\nEnter the starting crit chance: ");
scanf("%f", &fCrit);
printf("Enter the crit increase per attack: ");
scanf("%f", &fCritIncrease);
printf("Enter the number of trials: ");
scanf("%d", &nTrials);
while (nCounter <= nTrials)
{
fReturn = ChainFunction(fCrit, fCritIncrease);
fCritSum += fReturn;
if (nTrials < 20)
{
printf("Average crit chance on Trial %d was %f\n", nCounter, fReturn);
}
++nCounter;
}
printf("\nOverall crit percentage was %f\n", fCritSum/nTrials);
printf("Benefit from crit chain talent is %f\n\n", (fCritSum/nTrials) - fCrit);
return 0;
}
Retour en haut
![=]](https://lvlt.forum.bioware.com/public/style_emoticons/default/sideways.png)
Unfortunately, I'm not going to run the code through a compiler to reach the conclusion I already know.
.
