написан для анхоли, чтобы выяснить "че там с мастери то".
[Parse Damage]
This can be done with any combat log: for purpose of this post, I will be using
this one. DK I’m familiar with and who knows his stuff (although, he does make a mistake or two here, coughnodoublepotcough!), and Argaloth is an almost perfect boss for us to test our single target dps on; it’s essentially stationary, it has regular points at which you can soak magic damage for AMS RP, there’s no adds, there’s nothing to pull you off the boss, and so on. The only fault is that it’s relatively quick, which will likely invalidate its usefullness as people gear up, but for now, it still lasts long enough.
Anyways, this method is surprisingly easy, the flaw being RNG and specific fight factors can lead to misleading results, whereas the sim eliminates those two aspects. As long as your getting all your weights from the same specific parse, and said parse is as unremarkable as possible a fight, it works moderately well when the simulator isn’t an option and you don’t have the math expertise to start from scratch with the formulas.
Getting to it, you start by adding up all the damage affected by the specific stat in question. In this case, that means DC (612245), SS shadow (385857), BP (175109), and UB (48839). Because the patch obviously hasn’t gone live yet, you have to multiple all of these numbers by their various nerfs and buffs: DC (612245*0.9=551021), SS shadow (385857 *0.938*1.22=441559), BP (175109*0.93=162851), and UB (48839*0.9=43955). Remember, even though, yes, we have 8 base mastery for 20% shadow damage, you don’t want to factor that in here, as additional mastery isn’t multiplicative with that amount (i.e, you would do X * 1.21, not X * 1.2 * 1.01). Adding those numbers up, you get (551021 + 441559 + 162851 + 43955=) 1199386 damage. Thus an additional point of mastery rating would add you (1199386 * 0.025 * (1/180)=) 167 damage. Over the course of a 174 second fight, that’s 0.96 dps, and thus your weight.
Pretty good, right? Better than haste even!
The issue, however, is the length of the fight in question, which leads to higher than average Heroism/trinket/potion uptime, and thus is going to inflate any weight calculated from the parse in question. For true accuracy, you would have to repeat the process for each other stat – for purposes of this post, however, one is mainly concerned with how mastery faces up against crit. After all, if it’s worse, nothing else matters – we ignore the stat just as much as usual. If it’s better, then you would proceed to compare to the rest, but one thing at a time.
Crit is, of course, a trickier stat to calculate in such a manner. The simplest way to look at it would be to add up all the crittable damage in the parse which didn’t crit, multiply out any patch buff/nerfs (including the mastery one, in this case), then times by (parse crit rate per ability + 1/parse crit rate per ability/100), adding it all up, multiply by the crit meta (which we will, of course, be using), finished off by simply dividing by 180 and the fight duration. Fun, I know.
So: SS (497771 * .938 * .0110), DC (399609 * .9 * 1.2 * .0105), auto-attack (246555*0.0107), SS shadow (385857 * .73 * .938 * 1.22 * 1.2 * .0110), FeS (183772 * .938 * .0109), BP (141268 * .93 * 1.2 * .0108), FF (148985 * .93 * .0118), UB (48839 * .47 * .9 * 1.2 * 0.0105). Do note, that for SS shadow and UB, you have to multiply by the proportion of SS/DC’s non-crit to crit damage, as they’re not split. Anyways, adding this all up, you get 22037 * 1.03 = 22698. Divide by 180, and then divide by 174, and you wind up with 0.73 dps as your weight!
Now, of course, I did gloss over some factors. I ignored the UM nerf (as it would affect the two near equally, and it would greatly complicate things here). I assumed that the shadow part of SS is, indeed, intended to be affected by mastery, and that the ghoul still doesn’t scale with crit. I didn’t delve into the RC change, as it primarily affects ghoul damage (which doesn’t matter when using this method). Overall though, based off what we currently know, this is roughly accurate.
So, yes, Mastery does appear to be desirable – superior to crit/expertise, without a question, although it still falls short of haste (more on that at another time), and hence the title of the post! Yes, I was mistaken in my previous writing, and am more than happy to admit it. I have my own, personal spreadsheet I use to quickly test new changes and what have you, and I accidentally still had a part tweaked which was leading to the ghoul benefiting from crit rating (which would let it edge out mastery, for the record, but that’s neither here nor there, now). Hence my initial mis-estimation of the stat.
Тема: [Рыцари Смерти] Фрост в 4.0.6 (Прочитано 762871 раз)
