lets talk crit.

thefiresidecat

so I know they say up to a hundred percent there is no diminishing returns but is this really so? it seems like in most things in this game the rng percentage works on a per instance time not a percentage of the time as a whole. so every time I fire an encounter that's a x percent chance not over the session of a mob for instance total chance.


so if lets say I have a 75 percent change or a 90 percent chance. both of those are gonna be pretty darned near 100 percent in reality aren't they? I mean the dps missed from 75 or 80 or 90 or whatever is going to be statistically smaller then an actual ten or 20 percent missed crit chance. (I'm not a stats expert so by all means I may be full of crit but this is how I'm thinking lol)

so realistically where is the drop off in value for that stat in crit vs power? 100 seems nice but in reality are you hurting yourself when you do this vs stacking power instead at a point earlier than that?

kiraskytower

As with many things in statistics, saying 70% or 90% or whatever else assumes an infinite number of rolls of the dice. Sure, in a relatively short time span 90% crit might be the same as 100% crit (which is to say you may roll 90 or under for a whole fight) but it is also possible to have a fight where you manage to roll over 90% the entire fight and get no crits. The only thing 90% crit tells you is that over time 90% of your hits should be crits. So is 100% crit better than 90%? Of course it is in the long term, but you may not see a difference in any particular encounter.

niadan

The issue that plagues those that try to figure this out is that crit is entangled with other damage affecting mechanics. For example SS mages cannot proc SS without crit even though the crit is useless to the proc for increasing it's damage, since SS cannot crit (for the most part...).

Since in end game, it is relatively easy to max crit, it is just the way to go...similar to maxing ARP to 60%.

thefiresidecat

it's each time you fire an at will. each time you fire an encounter. the percent chance is each roll not over time. so if to trigger something you need one crit odds are more than in your favor at 90 percent. if you fire off three at wills one for sure will crit and most likely all three will crit. like when you're trying to upgrade something. the chance is not an over all chance of upgrade its each roll of the coin at (3% lets say) so every pres ward is a 3 percent chance to upgrade that thing to mythic. in this case every at will is a 90 percent chance.

looking at other things with a 90 percent chance to proc in this game.. it is almost ALWAYS going to proc. so it's not really a gamble. 75% is a little more of a gamble. 80 percent is pretty comfortable. 85 is looking pretty darned good. but 90 seems like it's probably near fail safe. I'm thinking you're only going to miss out on very rare extra damage here when you start to hit about 85-90% crit.

surely someone has played with this on act? this seems tailor made for act testing.

santralafax

It's hard to get a crisp 100 crit. I find I run with a 95-97% chance that ranges to 100. I can't find myself sacrificing the power it would take to keep >100% all the time. However, I am definitely sacrificing some (few) critical strikes in every boss fight.

theraxin

I'd suggest to count with the chance of not getting a critical attack, like out of 50 attack.

And

60% - 20 miss
80% - 10 miss
90% - 5 miss
95% - 2,5 miss
100% - 0 miss
On a note: 80% goes with every fifth attack missed. That not comfortable at all.

And this is pretty linear, you are spending equal amount of stat for the same damage output.

Too lazy for statistically count, but basically the bigger number goes with higher spread nominally. So while a 20% range of randomness in a 10 miss can go 8-12 miss, which can affect your DPS significantly, on 95% chance it's 2-3 miss spread what is not that noticeable.

So no, in that terms there's no diminishing in crit, and because in dungeons you are likely to give hundred of attacks, the crit chance will likely fall just around the said percentage.

Now, the problem is that there's a lot of criteria missing for a correct power vs crit decision.

like:
1, Your Power.
2, Your Crit Severity.
3, If you were running with 0% crit, how many percentage of your DPS could've critted, but not.

And just some tweaking with [1 + power/40000]*[base_damage]*[crit_chance]*[crit_sev]*[percentage_of_DPS_affected_by_crit]+[1 + power/40000]*[base_damage]*[100 - crit_chance] by adding 10% crit or 10% power*[1 - percentage_of_DPS_affected_by_crit].

Make the base damage like 1000 or 10000,.


4, Other (how likely that there is an AC DC in the group, or renegade CW...). This can't be counted, but can affect the decision.

tl;dr: Try to get it into at least 95%. You'll need that much later anyway.

thefiresidecat

theraxin#5169 said:

I'd suggest to count with the chance of not getting a critical attack, like out of 50 attack.

And

60% - 20 miss
80% - 10 miss
90% - 5 miss
95% - 2,5 miss
100% - 0 miss
On a note: 80% goes with every fifth attack missed. That not comfortable at all.

And this is pretty linear, you are spending equal amount of stat for the same damage output.

Too lazy for statistically count, but basically the bigger number goes with higher spread nominally. So while a 20% range of randomness in a 10 miss can go 8-12 miss, which can affect your DPS significantly, on 95% chance it's 2-3 miss spread what is not that noticeable.

So no, in that terms there's no diminishing in crit, and because in dungeons you are likely to give hundred of attacks, the crit chance will likely fall just around the said percentage.

Now, the problem is that there's a lot of criteria missing for a correct power vs crit decision.

like:
1, Your Power.
2, Your Crit Severity.
3, If you were running with 0% crit, how many percentage of your DPS could've critted, but not.

And just some tweaking with [1 + power/40000]*[base_damage]*[crit_chance]*[crit_sev]*[percentage_of_DPS_affected_by_crit]+[1 + power/40000]*[base_damage]*[100 - crit_chance] by adding 10% crit or 10% power*[1 - percentage_of_DPS_affected_by_crit].

Make the base damage like 1000 or 10000,.


4, Other (how likely that there is an AC DC in the group, or renegade CW...). This can't be counted, but can affect the decision.

tl;dr: Try to get it into at least 95%. You'll need that much later anyway.

testing it by counting is impossible for me. I'm a hr. I fire an atwill I often get two or three crits from each shot. I think it's roots critting. so I can't tell what has crit and what has not. ftr I've been at a hundred where it's occasionally proccing over a hundred. and currently I'm at 90 sometimes proccing to 96 or so. I was just seriously considering cutting it down a little more.

me personally with crit sev. I am running a t dread and my in game crit sev says 95% so not really sure how they add together. and my power procced is about 64k

theraxin

ACT (Advanced Combat tracker) is a good way to measure your critting or non critting skills.

The dread is not mentioned in the crit, but it's adds into it when an encounter is cast.

But, with these parameters, you should get on 100% crit (that might change with added information, atm I think all of your skills are critting).

That 2k power is insignificant in this level: The power multiplicator get's reduced by less than 2% (2,6 ->2,55), while you are having a 1,95 or 2,7 multiplicator upon every critting attack.

thefiresidecat

we don't have act on console that's why I posted here. lol good info tho ty.

thefabricant

If you are in a party with a power sharer, crit is the stat that gives the best returns and you will absolutely want it at 100%.

jonkoca

As an HR I try for around 100% during initial burst, with sudden precision and boons, and after 10 secs, settling down to around 90%. All the trash mobs die in that initial period, and for bosses, 90% with sporadic intervals of 100% is fine. Otherwise I stack power.

This is with a combat build, 150% critsev, 40-70k power and a t.feytouched. The usual pet with all the bells and whistles.

thefiresidecat

k ty guys. I put back my mount with crit instead of mount with power lol. 101%