SimulationCraft vs Ask Mr Robot - Krosus results

[OPness]

Hey guys so I normally use SimulationCraft for all of my simming, however, a guildy just started an Ask Mr Robot guild subscription. So I was wondering what the general consensus on the two were?

Here is why I am confused.

** I have the legendary pants and ring for arcane **

After using SimulationCraft, I am sitting at 30% crit, 21% haste, 21% mastery, and 8% vers. And I sim at ~710k.
H Krosus fight - https://www.warcraftlogs.com/reports/dV ... e&source=9

After using Ask Mr Robot's "best in bags" feature, it has me gain a lot more mastery than I thought I needed. I am sitting at 28% crit, 15% haste, 33% mastery, and 4% vers. I sim at ~690k
H Krosus fight - https://www.warcraftlogs.com/reports/tG ... e&source=4" target="_blank


So this is whats bothering me. I gained a lot more mastery and performed way better on krosus, even on a longer fight. Now to be fair I was more greedy with burning pitches and not moving all the way to the back (pug group), but I dont think that should make up for that much of a dps increase. I don't feel like my rotation was better on the second krosus fight but that could also be another factor in the increase.

I was under the assumption that by having the legendary pants mastery was not that much of an issue 15-20% should be fine then haste to ~22% and crit/vers at a 3:1 ratio - ish, which simulation craft agrees with. I guess I am just confused on why i performed much better.

[DrIMNLuv]

I could offer a few ideas after looking over your logs.
In the second fight you did the following:
Cast damaging spells more often (1.38 damaging casts per second vs 1.2 damaging casts per second). There's not much difference in %uptime for your haste buffs, so I would guess this is related to movement.
Blinked less (1 vs 6, that's 5 GCD of movement instead of blowing things up)
There's a burst during 2:10-2:50 of the fight that's just not in the first fight. I'm guessing it relates to a fight mechanic? But it's 30 seconds where you're basically not doing damage.
Had 54 artifact traits vs 51, only 2% difference in damage but it's still a factor to consider

I can't see/find your arcane charge uptime in warcraft logs, but that also represents a %damage increase that may differ between the two runs. However, I'm inclined to think movement in the first fight had a bigger impact than you think. Without a video we can't say. But you could always try the low mastery build again and be greedy about mechanics to see if that was the culprit. I'm only a mediocre mage, but I stare at data all day and those were the thoughts I had looking at your logs.

[OPness]

I agree with trying the fight out again, I was thinking about doing that before posting the thread but I didn't have enough time to find another heroic krosus group.

However back to the argument of SimCraft vs Ask Mr Robt, has anyone else experienced mixed results from the two? I'm still confused why it wanted me to almost double my mastery.

[Mage]

Are you really trying to draw conclusions from comparing 2 single fights?

[OPness]

Are you really trying to draw conclusions from comparing 2 single fights?

My original question...

Hey guys so I normally use SimulationCraft for all of my simming, however, a guildy just started an Ask Mr Robot guild subscription. So I was wondering what the general consensus on the two were?

I was more wondering why Ask Mr Robot would want me to increase my mastery by 15% and drop my haste even though that is not the highest sim I am getting with SimulationCraft.

The two fights were just my immediate testing with it. But thanks for the feedback!

[Mage]

I was more wondering why Ask Mr Robot would want me to increase my mastery by 15% and drop my haste even though that is not the highest sim I am getting with SimulationCraft.

Because they have different APLs.

[Daidai]

My interpretation for these two logs:
You were able to pull the same peak at startup without BL on the higher log. So you had 4 peaks vs. 3.
Simply RNG at the beginning, would be more equal if the peak would have been lower (as expected theoretically without BL) on the higher log.
You were also forced to blink during BL in the first phase on the lower log.

[DrIMNLuv]

So after doing some reading on Ask Mr Robot (AMR) I'll take a shot at addressing the differences between SimC and AMR. SimC and AMR calculate stat weights differently. When SimC calculates stat weights by adding 1500 to a stat, calculating the change in DPS, and the slope of that increase in DPS is your stat weight. It doesn't look at what would happen if you, for example, dropped 1500 haste to gain 1500 versatility. AMR tries to account for interactions in these stat changes and uses more than two points to calculate stat weights. Basically it uses more robust statistics to try to optimize stats than SimC.

I use SimC over Mr Robot because... well it's the one I have. But I might check out AMR now...

One thing to keep in mind is arcane mages have been successful using both high mastery and lower mastery builds. Spend more time trying them yourself and pick your preference.

[Kythos]

for your legendaries your first stats looks better to me. (30% crit, 21% haste, 21% mastery, and 8% vers)

#feelcraft but a handful of us in the #Arcane Dream discord think;

kilt allows for ~20% mastery
without kilt or Rhonin's or Gravity Spiral I suggest closer to 30% mastery.