Yet another simulation craft help thread

[cainito]

Ok, so simc is confusing me once again.

us.battle.net/ wow/ en/ character/ area-52/ Shambamheal/ simple (can't post links yet)

So I got my new trinket and boots, w00t! Throw them on, run simulation craft.

25000 interations, patchwork, all raid buffs, analyzing int, SP, hit, crit, mastery, and haste. Master/haste for reforget plots, 5000/500 for reforge and step amounts.

First test, I get to reforge.

Test one:

Int 6.97
SP 5.51
Hit 9.49
Crit 3.52
Haste 3.79
Mast 3.93


So, I gem and reforge according to that.

I run again with new chants and reforges.

AND NOW IT SAYS GO BACK TO HASTE?! YOU ROTTEN FUCKING FUCKER FUCK.

Int 7.02
SP 5.55
Hit 9.43
Crit 3.49
Haste 4.88
Mastery 3.65


Any help or explanation would be nice.

I didn't even close simulation craft between tests and I double checked to make sure I didn't change anything. At least i'm pretty sure I didn't change anything.

[Deleted]

Hello: I am the simulation craft helpboard. Please state your inquiry: It appears to be a human error of understanding...

Sorry

All joking aside, you need to remember the error margin. In your first test run you should consider mastery/haste to be more or less the same value, error margin give and take. Also, if possible, upload your reforge plots + provide armory link.

I'll provide a more exact answer if you can provide some more details

[cainito]

Hello: I am the simulation craft helpboard. Please state your inquiry: It appears to be a human error of understanding...

Sorry

All joking aside, you need to remember the error margin. In your first test run you should consider mastery/haste to be more or less the same value, error margin give and take. Also, if possible, upload your reforge plots + provide armory link.

I'll provide a more exact answer if you can provide some more details

http: // sdrv.ms/ 17GaFEP

there are the .html files for both of my tests, same gear is used in both, and same settings, or so I think

And I did provide armory in first post, second line.

[Caltiom]

Learn to understand what Scale Factors are. The whole problem is that you interpret them as something they are not.

Scale Factors are only valid linear approximations in a small area around the gear you simulate them with. If you change the gear based upon them, the scale factors will start to change, and you need to resimulate them. That doesn't mean you have to reforge for changing a few gems, but you shouldn't change 10'000 stat points all at once.
You'll often have two stats ( eg. haste & mastery ) compete for the first place. In this situation reforge plots are much better to find the sweet spot where their marginal values meet.

[Deleted]

http: // sdrv.ms/ 17GaFEP

there are the .html files for both of my tests, same gear is used in both, and same settings, or so I think

And I did provide armory in first post, second line.

Ah, below ten posts, no links, my bad.

The Sims are uploaded as videos for some reason, or so Skydrive claims o.O

EDIT: It's also worth mentioning that healing sims leaves much more to be interpreted in comparison to dps sims.

[cainito]

I am running DPS sim... hold on, filler post so I can get to 10 and post links..

BAMski

https://imageshack.com/i/f2z02cj
https://imageshack.com/i/mu4tq6j
https://imageshack.com/i/nm77jpj
https://imageshack.com/i/mk4qgyj

[MickyBIs]

As Caltiom said, scale factors and reforge plots are for different things, but are both based on your current gear.
Scale factors will tell you how much your DPS will improve by adding more of that stat, not how you should move your current stat points around. They're good for basing future gearing on, or determining upgrades.
Reforge plots will move your current stats around to show you the best DPS, so you need to look for the maximum on the graph. From your plots it looks like the graph1 reforges were optimal, but whatever you changed for graph2 made haste a very desirable stat. Even so, graph2 has a better set as the base DPS is higher.

If you want to test changes without doing it in game, some tools like askmrrobot can export a simulationcraft file for you, or you can modify the stat listings at the end of an in-application import; just remove the # at the beginning and it'll override the calculated amounts from gear with whatever value you put there.

[cainito]

Ok so, I run sim, get my results, the little icon/link for robot will put in those stat weights for me, and that is how I should reforge/gem, correct?

The thing is between the 2 graphs, the only thing I changed is the reforges and gems. The first test said mastery was higher, and clicked link for robot, had stat weights put in, what should be changed as far as gem/reforge, so I reforged the way robot said to go with stat weights from the sim. I then ran it again with updated gems/reforge and haste was a way higher priority.

What I'm not understanding is the difference in stat weights between the two tests, it is the same gear, just different gems/reforges. One said go mastery, the other says go haste. Or am I misunderstanding stat weights? I thought that if I click on the mr robot link in my results, stat weights would be put in, and that is the way I should gem/reforge to get the best dps?

[MickyBIs]

If you're changing gems pure for pure and mixed for mixed, it should have gone just as the first reforge plot showed, but if you changed out some pure int gems for mixed gems (which based on the weights you might have) you can end up with a lot more secondary stats. That can change the reforge priorities a bit, which is why I suggested resimming with the Mr Robot profile.

Either way, the second graph certainly has higher DPS, even with its current reforges. Since Mr Robot just tries to get the biggest number it can with the weights you give it, it can take a few sims or reforge plots to find the best balance of stats as there usually isn't one best overall stat that you'll be amazing with if you're fully stacked in it. Change the numbers a bit, move 5000 mastery to haste and see what it does. Can you manually reforge for that much? Is it better?

One important thing to remember about simulationcraft is that it uses an action list that may be very different from how you play, and it executes it perfectly. Something that makes a 5k DPS increase in simulationcraft might not be noticeable to you in game.