All good points. I guess I never really saw the value in it.
Don’t forget that target dummies are actually horrible for getting anything accurate for dps.
Besides some things not even working on them.
Thanks for this! I kinda figured it out when someone who was trying to give me pointers told me MoT doesn’t work on target dummies.
I figured out how to get Quick Sim to run the type of build I want to run on the target dummy (so I don’t have to use consumables to test). I’m getting 3.4-3.5k on it while Quick Sim is getting 4.1k. I’m trying to track down the difference but not terribly worried about it now, because it’s kinda making sense. I just got the Macabre Sheet music, which I have to say Quick Sim is far more generous about using than reality.
I believe I lose about 3-5% of possible dps to GCD/spell cast lag. Quick Sim has Flurry/IL/IL as 1.29s each. For me, they run 1.35.
There is one subtlety I’ve found to get another ~5% more dps that’s not obvious. You have to shatter frostbolt. I’m trying to figure out how to get Quick Sim to run different approaches/algorithms, but the problem is basically: if you cast frost bolt with brainfreeze up (maybe it pops just as you cast), but also 2x fingers of frost procs, your cast sequence should be:
IL/IL/FB/Flurry/IL/IL/FB --> and here there’s a 30% chance you have BF up while casting this FB
Otherwise, you’d cast: IL/IL/Flurry/IL/IL/FB/FB --> and then there’s a 30% chance you have BF up while casting the 2nd FB… (like the first case)
The difference is: you shatter the first FB in the first case, you don’t shatter it in the 2nd case. You’re going to cast it anyways (unless MoT procs)… so why not get the + 600-700 damage from the shatter?
Then the question becomes: if you proc 1 fof and 1 BF during cast of a FB… should you do IL/FB/Flurry/IL/IL/FB, or should you shatter your current FB cast and run over the fof on your first winters-chilled IL after Flurry. This one’s not obvious to me, because IL def. hits hard, but it forces you to cast another FB which may eat a BF that you can’t see. FB’s value is totally nonexistent if it can’t proc Flurry. FB by itself does a whopping 1400dps (2200/cast). The thing is, if it wasn’t for fof, then FB’s value comes from the fact that every 3.3 FB casts (for me, every 5.7s), it lets you shatter/Flurry/IL/IL for an average of another 18000 damage in ~4s - so =/=24600/9.7s = ~2500dps. If you don’t shatter the FB, I think you end up with an average more like 26800/11.4s=2350.
The rest of my dps comes from trinkets, frozen orb (which is like 8k dps, but has long c/d), and fof procs.
Aside from looking at all the timings, you should check damage % breakdown.
That’s the simplest way to figure out where you have rotation issues.
For a simplified example, if the sim tels you that ice lance should be 30% of your damage and it’s only 20% you know something is wrong there.
In my experience, if damage %'s are off on abilities by more than 1-2%, then there’s a significant rotation issue.
I haven’t tested frost on SL dummies, but I played it in pre-patch a lot. There was no issue matching sim dps pretty much exactly, on target dummies.
So I don’t buy your reasoning for being nearly 20% off on dps. As long as you’re simming properly, with all the same buffs and debuffs that you’re practicing with on the dummy.
The 20% was because I was trying to use Quick Sim default to match to a Target Dummy. Quick Sim has a target dummy option that drops raid buffs and matches pretty well. When I do that, Quick Sim gives me about 4100 dps, and I can kinda pull 3500/3600 doing the same things. With the target dummy option, the lack of MOT is handled correctly.
I am now running warcraftlogs, which you can run on training dummy tests.
I’m getting (looking at warcraftlogs vs. quick sim):
Me (warcraftlogs) : vs : Quick Sim
Ice Lance: 46.7% : 48.2%
Frostbolt: 18.1% : 14.2% (see note below)
Icicle: 15.5% : 13.8% (direct result of too heavy on FB)
Flurry: 14.6% : 15.6%, but note that hit/crit is hiding that I had 29 flurry when I should have been closer to 35.
Frozen Orb: 5.1% : 5.4%.
(Note: I’m probably losing flurrys because I’m trying not to waste fof… but I think you should just take your flurry if there’s only 1 fof. Note: edit, the Quick Sim numbers were using the wrong tinket, updated…)
Now that I have this kinda under control, I’m working on improving my testing and cleaning all this up. The toolset, in general, is applicable to my alts and some other stuff, so I’m on a good path to handling all this now.
1 Like
BF has Priority over FoF when only 1 FoF is present. If you get 2 FoF then you need to IL/FB/Flurry/IL/IL assuming you’re not under MoT. MoT changes things to a point that procs can over whelm you.