Critical Hit Chance
I saw this amazing open source google.doc on the blue garter forums which had A LOT of Critical Hit Data, so I gathered up my own Data-Points, then asked Viridiana, who I know has been gathering hundreds of Data Points, to add his to this google.doc… 354-760 Critical Hit Rating almost. Fantastic.
After adding over 400 data-points into my Solver, I created a Chart like this to see the Linear Sloping of CRT…
Giving me a Formulae for my data points CRT*0.0233-3.3009.
Now, I wanted to create a formula which scales CRT from 354, rather than 0, like what my slope is currently doing, so using my Excel Solver (which I used to calculate the Damage Model), I got a formula of:
Meaning, we have a base Crit Chance of 4.95% at 354 CRT.
You could also convert this into a whole number value, by changing a few numbers in the Formula to:
((CRT-354)*0.000232558)+0.49511233, giving you 0.0495.
Critical Hit Damage
Afterwards, I combined both Viridianas and a few other Data Points of Critical Hit Damage scaling.
Not as many data points, but it’s still enough for me to gauge a rough estimate of min/max values with and without crits.
The formulae I ended up with:
I noticed something here, that both Critical Hit Chance and Critical Hit Damage use the same 0.00023X value and so, there’s a huge possibility that the scaler for both Critical Chance and Critical Damage are exactly the same.
If this is the case, as I have more data points for Critical Hit Rating, this scaler would theoretically be more accurate than if I were solving straight for Critical Damage. If I then used my Critical Hit Rating Coefficient of 0.000232558 to solve Critical Damage:
Critical Hit Chance:
Critical Hit Damage: