I am probably not the first to make a spreadsheet to evaluate sharpshooter, but I made this for a friend and thought I might as well share it. The point is to answer the following questions: If I use sharpshooter and never stopped shooting at a target with infinite health (i.e. a spherical über chicken in a vacuum), then
1) what would be my effective crit chance?
2) by how much would my effective dps increase?
3) whats my real dps as compared to the insane number the game now displays?
4) what should I buy next, crit dmg or crit chance?
All you need to do is enter three numbers: crit chance, crit dmg and attack speed. The sheet then computes your effective damage as a multiple of the dps you'd have if you never ever critted. With and without sharpshooter. By comparing these numbers, one can say how much more dps you get from sharpshooter. In percent.
For the last question, I do not do three dimensional calculus, but simply compare the dps increase an additional 1% crit chance and an additional 10% crit dmg would give you.
I uploaded it to googledocs:
I think you need to download that as excel or open office to be able to enter your stats. Dont change any other cell except the three green ones.
The whole thing comfirms what is probably widely known anyway: Sharpshooter is only good if you have some increased crit dmg. And as you increase your attack speed and crit chance, it is worth less and less.
I.e. its a passive for low to medium gear. The spreadsheet just quantifies this.
For those interested to check my maths, here is what the columns do:
the number of shots fired without having had a crit before
the time that took in seconds, given the attack speed
the number of sharpshooter ticks. Thats just the time rounded down to the next integer
The next columns always come in threes. The first uses sharpshooter, the second does not, the third uses sharpshooter and assumes 1% more crit chance.
Chance to crit on that shot:
The chance to crit on the given shot assuming that there had not been a crit before. This starts at the input value in the 1st and 2nd column and 1% more in the third. It stays constant in the 2nd column and increases in the 1st and third by 0.03 times the number of ss ticks.
Chance of not having had a crit before, i.e. chance to even get that far:
This is the probability of all previous shots failing to crit. The 1st and 3rd column hit zero at some point because the chance to crit hits 1. The 2nd row converges to 0.
Chance of critting on exactly this shot for the first time:
The previous probability multiplied by the current chance to crit
Chance of at least one crit so far, including the current shot:
This simply sums up the probabilities in the previous three columns. Its just here to check things, as all three columns should converge to 1. And happily, they do.
Expected number of shots until first crit
The expected value of the number of shots needed to get the 1st crit is the infinite sum over all intergers, each multiplied with the probability of the 1st crit occurring on exactly that number.
This is the important bit. The numbers are totally meaningless taken on their own. The only thing that matters is the value the columns converge to if the number of shots converge to infinity. Infinity being 500 for all practical intents and purposes.
Thus, the only three values that are used in the results are the last in the three columns to the right. They give the average number of shots needed to get one crit. From that, I compute the effective crit chance and from that the effective dmg.
Edited by Xoth, 29 October 2012 - 06:21 PM.