I've been parsing a lot of combat logs as part of my weapons testing, and I've discovered that the amount of critical hits seems to be vastly higher than the level listed in the game's ship attack information.
It seems that if you have no points in Attack Vectors you will indeed get around 2.5% criticals. But if you max out Starship Attack Vectors your criticals don't increase to 4.5% as claimed by the game, instead they go up to more than 6%. This is without any points in any other skills related to space combat, and with only white MK 1 weapons and no consoles equipped.
From what I can tell, it doesn't seem like weapon power has any effect on the amount of criticals. As you get these much higher critical results whether your power is set to 50 or 125.
It appears to me that the Attack Vectors skill is giving a much larger bonus than the game would have you believe. There might also be some other bonus from higher weapon skill, because a higher level character is averaging closer to 7.6% criticals over the course of 376,000 shots fired. With that many individual results, more than luck has to be at play here.
Is there something I'm missing that can explain these extra criticals, or is critical percentage claimed by the in game text simply wrong?
My LG tac BoP runs a base of 4.5% crit chance and can go as high as 10-11% maximum for a few seconds. I have no idea what the average is though. Why would this be a big deal? With the high healing of STO, those criticals are important for destroying hard targets like cruisers, carriers and the like.
The issue could be with the interpretation of the numbers by your parser... and compound probability. No matter how ever many weapons you have equipped they will all have a 4.5% chance to land a crit. each shot.
However if you say fire 4 weapons your chance of landing one crit hit in that volley would be 16.82%
1 - ( ( 1 - x ) ^ y ) y=shots fired (or weapons firing) x=CritH
Of course with any chance for anything its still just that... chance. It wouldn't matter how large the sample you would never find an exact 4.5% in the numbers. There will always be a random element moving the numbers slightly up or down I would imagine.
It is also possible (and more likely) that your parser isn't doing anything strange with the numbers and acc is playing into the number. I have heard and don't remember ever seeing any 100% confirmation from any devs that acc over defensive values and such will lead not only to crit severity but also crit chance. If that is true perhaps you have found some degree of proof for that theory.
I don't really see how my parser can be doing something wrong, because it just counts how many shots hit, and how many of those were criticals. To get the percentage you just divide the critical count by the total count.
Even if there is a volley, so there is more chance of a critical among the shots fired, because there are more shots fired and thus more non-critical hits, it shouldn't change the percentage much. Also, I don't see any evidence of higher critical percentage with more weapons equipped, so it can't have anything to do with that.
I understand that you'll never get exactly 4.5%, but you don't get anywhere near 4.5%, it's not even close. But the odd thing is without putting points in attack vectors you do get very close to 2.5%.
As for the reason it matters. I just want to know what the default critical rate will be for the average person in the game. I'm assuming most people will put points in attack vectors. I'm trying to understand how the system works. Granted, you're probably right for the average player it makes little difference.
Another thing is this can impact what gear you choose to equip. +CritH adds 2.5% critical percentage. Let's assume that it really is 2.5%, that means if you were really getting 4.5% criticals, you'd be going from 4.5 to 7% with that weapon, an increase of 55%. But if you're really going from 7.5 to 10%, it's only an increase of 33%, which means it might be more worthwhile going with the 40% critical damage bonus from CritD equipment instead.