I have tested several talents, abilities and items that increase the damage of your critical strikes and they do not seem to be behaving consistently.
As an example I will take the Rogue talent Lethality which increases the critical strike damage modifier of some of your abilities by 30%. I take this to mean that crits will deal 2.3x damage instead of 2x and testing also confirms this.
Now the Mace Specialization talent for rogues says that it increases the damage of your critical strikes by 5%. Testing shows that Auto Attacks now will deal 2x * 1.05 damage on crit. Effectively making this equivalent to a 10% increase on your critical strike damage modifier.
When you combine these talents I would expect that the affected abilities of lethality will now deal 2.3 * 1.05 their base damage on crit. But testing shows critical strike damage higher than this.
I have some logs where I used Lethality (30% crit damage modifier), Mace spec (5% critical damage) and Relentless earthstorm diamond (3% critical damage).
I used a mace with a very small damage range (1-3) to get the most accurate results. The highest damage SS did 347 which should make an expected max crit of 3472.31.05*1.03=~863.2, but the found critical damage range was actually higher at 866-872 well above the expected value.
So how do these modifiers interact exactly?
PS: Can’t include links but I have logs uploaded on WCL showing these tests.
First question is have you accounted for AP scaling on your abilities, as well as weapon normalization? Where all the other values the same (same target, target level, armor value) etc. Your above information does not give enough information
Yes all tests were done on the same target dummy with the same debuffs (none at all). I’m not sure what you mean with mentioning the AP scaling and normalization since they are not relevant here.
Weapon strikes use a weapon normalization value, which in effect changes so that the raw damage of a weapon has less impact, to nullify lower dps weapons with a slow speed perform better then higher dps weapons with a fast speed. This weapon speed changes the end value of your hits (but white and yellow)
This is multiplied by AP scaling abilities. Some abilities scale with attack power, the more attack power the more damage they do. this is then adjusted with weapon normalization values.
These aspects affect your testing.
Try to get a 2.7 main hand, 2.4 offhand, or 1.7 speed dagger and test the values with those speeds. Are they correct then, then its probably due to the above mentioned weapon normalization.
Are they incorrect? Submit a bug report and include the weapon speed of your weapons (preferable the name of the weapon tested as well, as well as an accurate as possible time index.)
I don’t think we are communicating correctly. I’ll give a mathematical example
Weapon speed is 2.5 instead of 2.7 or 2.9 instead of 2.7
Example faster weapon then 2.7
Sinister strike damage 347 * 2.3 * 1.05 * (2.7/2.5) = 905 (this gets reduced by armor)
Example slower weapon then 2.7
Sinister strike damage 347 * 2.3 * 1.05 * (2.7/2.9) = 780 (this gets reduced by armor)
Additionally your base attack that gets also gets different ap scaling values percentage wise, slower weapons gain slightly more ap scaling to make up for normalization again. About 0.8% per 0.1 second speed difference
All this is done to try to make most weapons regardless of speed relevant if their weapon dps is higher then your previous weapon.
So in conclusion it might verry well be the correct numbers that your seeing, although you should still make a ticket in my opinion to make them notice and double check it