I have been looking at the Fury Warrior sim on both here and wowsims at github.
AMR prioritizes crit gems over STR, but other sims says pure STR gems are the best. All of the top parsers use full STR gems as well. Is there a way to get AMR to prioritize STR gems when you sim?
Just looked at the templet fury warrior.
The difference between crit and strength gems is extremely small – one different assumption in the model being used can push it one way or the other.
If you use this other tool and put in all crit vs. all strength gems, how much of a difference does it predict?
Have you confirmed that all of the assumptions used on AMR and in the simulation are the same? All buffs/debuffs/consumables are the same, fight length, etc.?
Just find it strange that all other sources except AMR values str higher then crit in all sims. If you look at all logs, noone is gemming crit.
If you compare AMR to WOTLK Warrior Simulator
AMR Bis Templet Fury warrior p2
snap id: d910d44863614abba1e11d4e503daf90
Wowsim: Bis Templet Fury warrior p2
Wowsims bis in AMR:
91f3a5603c1b442e8c52f6572110e88a
AMR bis in Wowsims
Its a massiv 700 dps diffrence
so either Stacking crit is a secret tactic that noone has figured out yet, or something is off
I can take a look why the two models are coming up with such a big difference, but on the surface, it doesn’t seem to have much to do with gemming crit vs strength. If you look, the setup that this other tool thinks is best actually has less strength than the one we have with mostly crit gems.
So it must be something about the other chosen items that this other site thinks is really good. I’d have to dig into it further to see what is really causing the difference. Locking in the set items doesn’t make a huge difference, the darkmoon trinket isn’t that big of a difference… so I’ll have to look further.
One thing that I could use some help verifying… it seems that some of the rating conversions in WotLK Classic are not the same as they were back in original WotLK.
For example, they changed resilience (which another user pointed out), so I re-extracted the combat rating conversions and sure enough, it takes more resilience than it did back in the day to get 1% crit reduction.
The data file is a little ambiguous… but I believe armor penetration has changed as well. Has anyone tested this recently?
I bring this up because I notice that this simulator you linked is using the outdated resilience rating conversion. It wants quite a bit more armor penetration as well, partly due to using a different conversion value (13.99, which was the old value from original WotLK, but I think it is 15.395300 on the live WotLK servers now).
643/13,99= 45,96
pre 3.2.0 that was 12.31 before it was changed to 13.99. Where is the 15,395 value from?
From the extracted combat ratings… the column is unlabeled, but it is the only value that seems close to what armor pen should be. But if the game is still showing a value of ~13.99 in the tooltips… then I guess they don’t put that rating conversion in that file for some reason, so perhaps that column corresponds to something else.
The file is combatratings.txt, can be extracted from any build, and is usually where one finds rating conversions.
They did change resilience though… someone should tell that simulator to update it to the new value.
Still doesnt answer the question why AMR wants to go full crit gem, when all theory crafters and top parses go str and more agi
I’ll look into it – we are finishing up 10.1 this week, so the next site update will be monday or tuesday next week.
Agility is never a better stat than crit for DPS warriors, it would be incidental because certain items with agility have other desirable stats. Agi converts to crit at a worse rate than crit rating and doesn’t give AP for warriors.
